Supervised Link Prediction Research Articles

Random walk with restart on multilayer networks: from node prioritisation to supervised link prediction and beyond

BackgroundBiological networks have proven invaluable ability for representing biological knowledge. Multilayer networks, which gather different types of nodes and edges in multiplex, heterogeneous and bipartite networks, provide a natural way to integrate diverse and multi-scale data sources into a common framework. Recently, we developed MultiXrank, a Random Walk with Restart algorithm able to explore such multilayer networks. MultiXrank outputs scores reflecting the proximity between an initial set of seed node(s) and all the other nodes in the multilayer network. We illustrate here the versatility of bioinformatics tasks that can be performed using MultiXrank.ResultsWe first show that MultiXrank can be used to prioritise genes and drugs of interest by exploring multilayer networks containing interactions between genes, drugs, and diseases. In a second study, we illustrate how MultiXrank scores can also be used in a supervised strategy to train a binary classifier to predict gene-disease associations. The classifier performance are validated using outdated and novel gene-disease association for training and evaluation, respectively. Finally, we show that MultiXrank scores can be used to compute diffusion profiles and use them as disease signatures. We computed the diffusion profiles of more than 100 immune diseases using a multilayer network that includes cell-type specific genomic information. The clustering of the immune disease diffusion profiles reveals shared shared phenotypic characteristics.ConclusionOverall, we illustrate here diverse applications of MultiXrank to showcase its versatility. We expect that this can lead to further and broader bioinformatics applications.

Embedding-based Method for the Supervised Link Prediction in Social Networks

In recent years, social network analysis has received a lot of interest. Link prediction is an important area of research in this field that uses information from current networks to predict the likely links that will emerge in the future. It has attracted considerable attention from interdisciplinary research communities due to its ubiquitous applications in biological networks, computer science, transportation networks, bioinformatics, telecommunication networks, and so on. Currently, supervised machine learning is one of the critical techniques in the link prediction task. Several algorithms have been developed by many authors to predict the future link in the network, but there is still scope to improve the previous approaches. In the supervised link prediction process, feature selection is a crucial step. Most existing algorithms use one type of similarity-based feature to represent data, which is not well described due to the large scale and heterogeneity of social networks. One of the newest techniques for link prediction is embedding methods, which are used to preparing the feature vector for each the nonexisting links in the network. In this paper, we introduce a novel approach to supervised link prediction based on feature embedding methods in order to achieve better performance. Our contribution considers a set of embedding methods as the feature vector for training the machine learning classifiers. The main focus of this work is to investigate the relevance of different feature embedding methods to improve the performance of the supervised link prediction models. The experimental results on some real-world temporal networks revealed satisfactory results, which encourage us for further analysis. Moreover, the use of feature embedding methods will provide better performance in this regard.

Effective Indexes and Classification Algorithms for Supervised Link Prediction Approach to Anticipating Technology Convergence: A Comparative Study

This article conducts a comparative analysis to investigate the effects of different classification algorithms and structural proximity indexes on the performance of the supervised link prediction approach to anticipating technology convergence at different forecast horizons. For this, we identify relationships between technologies of interest for different time periods and compute 10 structural proximity indexes among unconnected technologies at each period. We develop a set of classification models that identify potential convergence among unconnected technologies where each model is configured differently by a classification algorithm and a combination of the proximity indexes. We compare the performance of the classification models to investigate effective combinations of classification algorithms and proximity indexes at different forecast horizons. The empirical analysis on Wikipedia articles about artificial intelligence technology indicates that random forest outperforms others in short-term forecasting while support vector machine outperforms others in mid-term forecasting. We also identify structural proximity indexes that produce higher performance when combined with the most effective algorithm at each forecast horizon. The results of this article are expected to offer guidelines for choosing classification algorithms and indexes when applying the supervised link prediction approach in anticipating technology convergence.

Evolutionary Features for Dynamic Link Prediction in Social Networks

One of the inherent characteristics of dynamic networks is the evolutionary nature of their constituents (i.e., actors and links). As a time-evolving model, the link prediction mechanism in dynamic networks can successfully capture the underlying growth mechanisms of social networks. Mining the temporal patterns of dynamic networks has led researchers to utilise dynamic information for dynamic link prediction. Despite several methodological improvements in dynamic link prediction, temporal variations of actor-level network structure and neighbourhood information have drawn little attention from the network science community. Evolutionary aspects of network positional changes and associated neighbourhoods, attributed to non-connected actor pairs, may suitably be used for predicting the possibility of their future associations. In this study, we attempted to build dynamic similarity metrics by considering temporal similarity and correlation between different actor-level evolutionary information of non-connected actor pairs. These metrics then worked as dynamic features in the supervised link prediction model, and performances were compared against static similarity metrics (e.g., AdamicAdar). Improved performance is achieved by the metrics considered in this study, representing them as prospective candidates for dynamic link prediction tasks and to help understand the underlying evolutionary mechanism.

Negative link prediction to reduce dropout in Massive Open Online Courses.

In recent years, the rapid growth of Massive Open Online Courses (MOOCs) has attracted much attention for related research. Besides, one of the main challenges in MOOCs is the high dropout or low completion rate. Early dropout prediction algorithms aim the educational institutes to retain the students for the related course. There are several methods for identification of the resigning students. These methods are often based on supervised machine learning, and require student activity records to train and create a prediction model based on the features extracted from the raw data. The performance of graph-based algorithms in various applications to discover the strong or weak relationships between entities using limited data encouraged us to turn to these algorithms for this problem. Objective of this paper is proposing a novel method with low complexity, negative link prediction algorithm, for the first time, utilizing only network topological data for dropout prediction. The idea is based on the assumption that entities with similar network structures are more likely to establish or remove a relation. Therefore, we first convert the data into a graph, mapping entities (students and courses) to nodes and relationships (enrollment data) to links. Then we use graph-based algorithms to predict students' dropout, utilizing just enrollment data. The experimental results demonstrate that the proposed method achieves significant performance compared to baseline ones. However, we test the supervised link prediction idea, and show the competitive and promising results in this case as well. Finally, we present important future research directions to improve the results.

Heterogeneous graph construction and HinSAGE learning from electronic medical records

Graph representation learning is a method for introducing how to effectively construct and learn patient embeddings using electronic medical records. Adapting the integration will support and advance the previous methods to predict the prognosis of patients in network models. This study aims to address the challenge of implementing a complex and highly heterogeneous dataset, including the following: (1) demonstrating how to build a multi-attributed and multi-relational graph model (2) and applying a downstream disease prediction task of a patient’s prognosis using the HinSAGE algorithm. We present a bipartite graph schema and a graph database construction in detail. The first constructed graph database illustrates a query of a predictive network that provides analytical insights using a graph representation of a patient’s journey. Moreover, we demonstrate an alternative bipartite model where we apply the model to the HinSAGE to perform the link prediction task for predicting the event occurrence. Consequently, the performance evaluation indicated that our heterogeneous graph model was successfully predicted as a baseline model. Overall, our graph database successfully demonstrated efficient real-time query performance and showed HinSAGE implementation to predict cardiovascular disease event outcomes on supervised link prediction learning.

A Supervised Link Prediction Method Using Optimized Vertex Collocation Profile

Classical link prediction methods mainly utilize vertex information and topological structure to predict missing links in networks. However, accessing vertex information in real-world networks, such as social networks, is still challenging. Moreover, link prediction methods based on topological structure are usually heuristic, and mainly consider common neighbors, vertex degrees and paths, which cannot fully represent the topology context. In recent years, network embedding models have shown efficiency for link prediction, but they lack interpretability. To address these issues, this paper proposes a novel link prediction method based on an optimized vertex collocation profile (OVCP). First, the 7-subgraph topology was proposed to represent the topology context of vertexes. Second, any 7-subgraph can be converted into a unique address by OVCP, and then we obtained the interpretable feature vectors of vertexes. Third, the classification model with OVCP features was used to predict links, and the overlapping community detection algorithm was employed to divide a network into multiple small communities, which can greatly reduce the complexity of our method. Experimental results demonstrate that the proposed method can achieve a promising performance compared with traditional link prediction methods, and has better interpretability than network-embedding-based methods.

Network Embedding For Link Prediction in Bipartite Networks

Many social networks have a bipartite nature. Link prediction in social networks has been the focus of interest for many researchers recently. Network embedding, which maps each node in the network to a low-dimensional feature vector is used to solve many problems. The aim of this study is to investigate how network embedding enhance the link prediction performance in bipartite networks. A network embedding and a supervised learning based link prediction model has been presented for bipartite networks. The input of the supervised learning model is learned embedding vectors of node pairs obtained from network embedding method. The target feature of prediction is a binary label indicating the existence or absence of a link between these node pairs. Ensemble learning algorithms have been applied for supervised link prediction. The experiments performed on two bipartite social networks built from public datasets led promising results with 0.939 and 0.974 AUC values. Random Forest models trained with embedding vectors obtained from BiNE method achieved the highest performances.

Transfer learning based attack detection for wireless communication networks

AbstractWireless communication network attack detection is important in the field of wireless communication network data mining. Transfer learning is a powerful tool to solve the problem of link prediction in unlabeled wireless networks. In the feature mapping process of transfer learning, the loss of source domain information is very serious, which leads to the loss of knowledge in inverse mapping. To obtain more specific and complete information in the source domain, a link prediction method based on the distribution function fitting method of domain orthogonal term family is proposed. The key samples in the source domain are used to construct a family of orthogonal polynomials, and the least square approximation of the distribution function is obtained by using the family of orthogonal polynomials. The feature vector set of the key samples is extracted from the source domain, and the pseudo label is assigned to the target training set, so that the link classifier can complete the supervised link prediction task. In addition, an advanced performance of TLAD is obtained by comparing with other state‐of‐the‐art models in practice.

Supervised link prediction using structured‐based feature extraction in social network

SummarySocial network analysis (SNA) has attracted a lot of attention in several domains in the past decades. It can be of 2‐folds: one is content‐based, and another one is structured‐based analysis. Link prediction is one of the emerging research problems, which comes under structured‐based analysis that deals with predicting the missing link, which is likely to appear in the future. In this article, the supervised machine learning techniques have been implemented to predict the possibilities of establishing the links in future. The major contribution in this article lies in feature construction from the topological structure of the network. Several structured‐based similarity measures have been considered for preparing the feature vector for each nonexisting links in the network. The performance of the proposed algorithm has been extensively validated by comparing with other link prediction algorithms using both real‐world and synthetic data sets.

Supervised link prediction in multiplex networks

In recent years, multiplex networks have been introduced to describe real complex systems, where the same group of entities make different types of interaction. In a multiplex network, each layer expresses one distinct type of interaction. Link prediction is a research hotspot in complex network analysis. A large number of link prediction methods have been proposed, but only a few were designed for multiplex networks. In this paper, we focus on the link prediction problem in multiplex networks. In our opinion, an approach in which link prediction is performed by simultaneously considering the information from all layers is advisable, because the formation of links in one layer can be affected by links of the same node pairs in other layers. A supervised method is proposed in this study to implement link prediction in multiplex networks, which regards link prediction as a binary classification problem. In the proposed method, a classification model is fed by a set of elaborate structural features of node pairs that are extracted from all layers. Extensive experiments are conducted on six networks to analyze the effectiveness of the proposed method. The results demonstrate that the proposed method outperforms the compared methods significantly.

Bursting the Filter Bubble: Fairness-Aware Network Link Prediction

Link prediction is an important task in online social networking as it can be used to infer new or previously unknown relationships of a network. However, due to the homophily principle, current algorithms are susceptible to promoting links that may lead to increase segregation of the network—an effect known as filter bubble. In this study, we examine the filter bubble problem from the perspective of algorithm fairness and introduce a dyadic-level fairness criterion based on network modularity measure. We show how the criterion can be utilized as a postprocessing step to generate more heterogeneous links in order to overcome the filter bubble problem. In addition, we also present a novel framework that combines adversarial network representation learning with supervised link prediction to alleviate the filter bubble problem. Experimental results conducted on several real-world datasets showed the effectiveness of the proposed methods compared to other baseline approaches, which include conventional link prediction and fairness-aware methods for i.i.d data.

A hotel recommendation system based on customer location: a link prediction approach

Hotel recommendation is one of the most used application areas in recommendation systems. So far, many hotel recommendation systems have been proposed. Most of these systems are based collaborative filtering, content-based filtering, and association rule methods and employ the features of hotel, the ratings given by user, online reviews and comments in social network about the related hotel as data. However, due to the difficulty of processing the data used, the performance rates and speeds of these methods are relatively slow. As a solution to these problems in this paper, we propose a novel hotel recommendation system based on link prediction method. For this purpose, a customer-hotel bipartite network was first constructed and the relationship information in this network was used as data. Then, a supervised link prediction method that consider customers’ location was presented. To the best of our knowledge, this is the first study that recommends hotel by using link prediction method. The experimental results conducted on data crawled from TripAdvisor.com demonstrate that the proposed method captures an accuracy of 89.5% and outperforms the other recent related algorithms.

A pattern based supervised link prediction in directed complex networks

Link prediction is one of the most interesting tasks in complex network analysis. Numerous recently published link prediction methods have focused on utilizing network models close to real networks to improve performance of link prediction. Directed, temporal, weighted and heterogeneous network models are some examples of the favored network models. Most published link prediction metrics cannot take into account the effect of links directions on link formation. In this study, we propose a pattern based supervised link prediction approach to improve link prediction accuracy of Triad Closeness (TC) metric in directed complex networks. The proposed pattern based link prediction metric is compared with TC metric and the state-of-the-art link prediction metrics to evaluate the effectiveness of the proposed metric. Experimental results in two citation networks show that the proposed metric improves remarkably link prediction accuracy of TC metric and obtains the highest link prediction performance compared to the state-of-the-art link prediction metrics.

Predicting Citation Count of Scientists as a Link Prediction Problem.

The studies dealing with the problem of predicting scientific impacts in the scientific world mostly focus on predicting citation count of papers (PCCP). However, in the literature, only a little bit of research has been conducted on estimating the future influence of scientists individually. Estimating the impact of scientists individually is a worthwhile task for the following scientific research and cooperatives. From this point of view, a new supervised link prediction method is proposed to predict the citation count of scientists (PCCS). Many PCCP studies employ document-based attributes, such as titles, abstracts, and keywords of papers; institutions of scientists; impact factors of publishers; etc. and they do not take advantage of any topological features of complex networks formed with citations among papers. However, citation networks include valuable features for PCCP and PCCS. Therefore, we formulate the problem of PCCS as a link prediction problem in directed, weighted, and temporal citation networks. The proposed approach predicts not only links but also its weights. Our supervised link prediction method is tested on two citation networks in Experiment 1. The results of Experiment 1 confirm that our method achieves promising performances when considering prediction links with its weights are addressed for the first time in terms of link prediction in directed, weighted, and temporal networks. In Experiment 2, the performance of the proposed link prediction metric and five well-known link prediction metrics are compared in terms of prediction new links in complex networks. The results of Experiment 2 demonstrate that the proposed link prediction metric outperforms all baseline link prediction metrics.

Predicting potential drug-drug interactions on topological and semantic similarity features using statistical learning.

Drug-drug interaction (DDI) is a change in the effect of a drug when patient takes another drug. Characterizing DDIs is extremely important to avoid potential adverse drug reactions. We represent DDIs as a complex network in which nodes refer to drugs and links refer to their potential interactions. Recently, the problem of link prediction has attracted much consideration in scientific community. We represent the process of link prediction as a binary classification task on networks of potential DDIs. We use link prediction techniques for predicting unknown interactions between drugs in five arbitrary chosen large-scale DDI databases, namely DrugBank, KEGG, NDF-RT, SemMedDB, and Twosides. We estimated the performance of link prediction using a series of experiments on DDI networks. We performed link prediction using unsupervised and supervised approach including classification tree, k-nearest neighbors, support vector machine, random forest, and gradient boosting machine classifiers based on topological and semantic similarity features. Supervised approach clearly outperforms unsupervised approach. The Twosides network gained the best prediction performance regarding the area under the precision-recall curve (0.93 for both random forests and gradient boosting machine). The applied methodology can be used as a tool to help researchers to identify potential DDIs. The supervised link prediction approach proved to be promising for potential DDIs prediction and may facilitate the identification of potential DDIs in clinical research.

Multi-kernel one class link prediction in heterogeneous complex networks

The heterogeneity of a network causes major challenges for link prediction in heterogeneous complex networks. To deal with this problem, supervised link prediction could be applied to integrate het...

A link prediction method for heterogeneous networks based on BP neural network

Most real-world systems, composed of different types of objects connected via many interconnections, can be abstracted as various complex heterogeneous networks. Link prediction for heterogeneous networks is of great significance for mining missing links and reconfiguring networks according to observed information, with considerable applications in, for example, friend and location recommendations and disease–gene candidate detection. In this paper, we put forward a novel integrated framework, called MPBP (Meta-Path feature-based BP neural network model), to predict multiple types of links for heterogeneous networks. More specifically, the concept of meta-path is introduced, followed by the extraction of meta-path features for heterogeneous networks. Next, based on the extracted meta-path features, a supervised link prediction model is built with a three-layer BP neural network. Then, the solution algorithm of the proposed link prediction model is put forward to obtain predicted results by iteratively training the network. Last, numerical experiments on the dataset of examples of a gene–disease network and a combat network are conducted to verify the effectiveness and feasibility of the proposed MPBP. It shows that the MPBP with very good performance is superior to the baseline methods.

Link Prediction on a Network of Co-occurring MeSH Terms: Towards Literature-based Discovery.

Literature-based discovery (LBD) is a text mining methodology for automatically generating research hypotheses from existing knowledge. We mimic the process of LBD as a classification problem on a graph of MeSH terms. We employ unsupervised and supervised link prediction methods for predicting previously unknown connections between biomedical concepts. We evaluate the effectiveness of link prediction through a series of experiments using a MeSH network that contains the history of link formation between biomedical concepts. We performed link prediction using proximity measures, such as common neighbor (CN), Jaccard coefficient (JC), Adamic / Adar index (AA) and preferential attachment (PA). Our approach relies on the assumption that similar nodes are more likely to establish a link in the future. Applying an unsupervised approach, the AA measure achieved the best performance in terms of area under the ROC curve (AUC = 0.76), followed by CN, JC, and PA. In a supervised approach, we evaluate whether proximity measures can be combined to define a model of link formation across all four predictors. We applied various classifiers, including decision trees, k-nearest neighbors, logistic regression, multilayer perceptron, naïve Bayes, and random forests. Random forest classifier accomplishes the best performance (AUC = 0.87). The link prediction approach proved to be effective for LBD processing. Supervised statistical learning approaches clearly outperform an unsupervised approach to link prediction.

A new non-parametric feature learning for supervised link prediction

Link prediction is an important task for analysing relational data such as the friendship relation on a social networking website that also has applications in other domains like, information retrieval, bioinformatics and e-commerce. The problem of link prediction is to predict the existence or absence of edges between vertices of a network. In this paper, we present a novel non-parametric latent feature relational model based on distance dependent Indian buffet process (DDIBP), by which we can utilise the information of topological structure of the network such as shortest path and connectivity of the nodes and incorporate them into the proposed Bayesian Non-parametric latent feature model which can automatically infer the unknown latent feature dimension. We also develop an efficient MCMC algorithm to compute the posterior distribution of the hidden variables with a highly nonlinear link likelihood function. Experimental results on four real datasets demonstrate the superiority of the proposed method over other latent feature models for link prediction problem.